Known prior art gain control circuits have been implemented exclusively with analog devices, by combining both analog circuits and digital circuits or with only digital electronics. Known prior art gain control circuits detect a single characteristic parameter of the input signal which is to be adjusted. The selected parameter may include the main valve, effective value or peak value of the signal.
Where it is desirable, because of downstream processing, to use a gain control circuit which includes digital elements problems have been encountered in processing analog input signals having large dynamic ranges. The problems arise out of the fact that the digital circuitry, which might be implemented in part as a microprocessor, is usually organized to have a predetermined digital word length.
If the dynamic range of the analog signal is too large, it will at times exceed the predetermined digital word length. When this happens, it is necessary to resort to multiple word arithmetic which can be slower and more complicated than single word arithmetic.
In some applications, the input analog signal to be processed may lie within a relatively narrow, predetermined frequency range and have an overall known wave form the dynamic range of which varies. At times, the signal which is intended to be processed may be distorted due to noise superimposed thereon.
If the amplitude of the noise exceeds that of the signal to be processed, it will have an undesired effect on the operation of the automatic gain control circuitry. Hence, in addition to being able to deal with an input signal of a varying dynamic range, it is also desirable to be able to filter the signal prior to carrying out the gain control-related processing.